As we look forward to continued growth and technological advances in the year 2016, we wanted to take a quick look back at some of our favorite articles and news stories about big data and analytics published throughout 2015. From advances in self-driving cars to increases in farming efficiency, analytics has played a role in many different kinds of achievements.

We’ve divided these articles into two categories: how data science works and how analytics is used to solve various problems.

How Data Science Works

VisualIntro_r2d3_Screenshot
1. A Visual Introduction to Machine Learning   |  r2d3, July 2015

An easy-to-understand, highly visual explanation of basic machine learning concepts and terms using a data set of homes in San Francisco and New York. (This was one of the most popular articles we shared all year!) Read it >>


HowMachinesLearn_HBR_Screenshot
2. How Machines Learn (And You Win)   |  Harvard Business Review with Randy Olsen & r2d3, November 2015

This article gives a simplified explanation of machine learning for those that may not be tech- or statistics-savvy, and shows an example of how a decision tree can be used to identify cable customers most likely to churn. Read it >>


3. If Algorithms Know All, How Much Should Humans Help?   |  Steve Lohr, New York Times, April 6, 2015

Mathematical models are often created using vast amounts of information—more than any human could analyze and comprehend. However, there is a need for transparency into the algorithms, especially in certain fields. This article explores the role of human intervention in the development of these algorithms. Read it >>


Analytics in Action

4. Quantifying the Impact of Marketing Analytics   |  Matt Ariker, Alejandro Diaz, Christine Moorman, & Mike Westover, Harvard Business Review, November 5, 2015

Surveys show that many companies are currently spending a portion of their marketing budget on analytics, and most plan to spend even more in the future. This article addresses two major questions:

  1. Do marketing analytics improve profits or ROI?; and
  2. Are companies using analytics effectively, and how does one quantify the effect?

The results of surveys show that companies who devote themselves to activating insights rather than simply generating them are more likely to experience success from their marketing analytics efforts. It is also suggested that effective communication is key, and the best way fill the gap is to hire an analytics translator. The last piece of advice is to start small; only tackle 1-2 new projects at a time. Read it >>


cow milking machines
5. How RFID Delivers Big Data On Cows And Milk Production   |  Brian T. Horowitz, TechCrunch, November 3, 2015

When thinking about uses for analytics, dairy farming is probably not the first that comes to mind. However, the dairy industry is now using RFID and sensors to collect data on its cows with an ultimate goal of maximizing milk production and minimizing the number of cows needed to keep up with milk demand. The data is used to monitor the diets, blood profiles, and overall health of these animals, and to alert farmers when additional intervention is needed. Read it >>


Best Roadtrip Map
6. Computing the optimal road trip across the U.S.   |  Randy Olsen, March 8, 2015

Analytics aren’t solely used for business purposes; some people like to use algorithms to answer questions purely for enjoyment; in this case, researcher Randy Olsen uses an algorithm to calculate the optimal road trip across the U.S., being sure to factor in stops at national parks, landmarks, monuments, and historic sites. Read it >>


John Lynn Screenshot
7. Podcast: The Guy Who Predicts Whether A Movie Will Bomb, Months Before It’s Made   |  Jody Avirgan with Andy Greenland, FiveThirtyEight, September 17, 2015

How does a movie studio know whether a movie is worth making…or if it will be a dud? In the past, executives would be forced to rely on educated guesses. Not anymore! Josh Lynn of Piedmont Media Research has created an algorithm that takes data on the basic plot points and cast, and can predict how successful a movie will be at the box office. He discusses the algorithm and some interesting insights he has discovered. Read it >>


DJ Patil, Chief Data Scientist of the U.S.
8. Obama’s Chief Data Scientist Reveals How the Government Uses Big Data   |  Tessa Berenson, Time, September 26, 2015

The nation’s first Chief Data Scientist, DJ Patel, talks about the effort to establish Data.gov, a repository of over 190,000 publicly available datasets, and discusses how analytics can be used to improve the lives of United States citizens. Read it >>


Mixtape from Quartz Spotify article
9. The magic that makes Spotify’s Discover Weekly playlists so damn good   |  Adam Pasick, Quartz, December 21, 2015

Many people have fond memories of receiving specially curated collections of music from friends. Listening to those painstakingly chosen tracks often made the listener think, “This is my new favorite song,” or “Wow, this song is exactly how I am feeling!”

Using a combination of algorithms that analyze user-generated playlists and listener preferences and with techniques such as collaborative filtering, natural language processing, and deep learning, Spotify has succeeded in creating weekly playlists for users that bring to mind the carefully crafted mixtapes of the past. Read it >>


self-driving car image via The Economist
10. How Data Science is Driving the Driverless Car   |  Hannah Augur, Dataconomy.com, December 21, 2015

Self-driving cars are no longer simply an idea from science fiction; a number of companies are hard at work developing the technology allows cars to drive autonomously. This article gives a brief overview of the hardware that allows a driverless car to “see”, and the software that allows the car to “think” and make decisions. Data plays a huge part in both of these processes; the hardware generates massive amounts of data, and the software relies on complicated algorithms that learn from data collected on previous drives. Read it >>